Pianta (2005)


Published on

Published in: Technology, Economy & Finance
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Pianta (2005)

  1. 1. J. Fagerberg, D. Mowery and R. Nelson (eds), Handbook of Innovation, Oxford University Press (forth.) Chapter 22 Innovation and Employment Mario Pianta July 2003IntroductionThe relationship between innovation and employment is a complex one and has long been atopical issue in economic theory1. The classical question "does technology creates or destroysjobs?" has to be asked in well defined contexts. Considering the disequilibrating nature oftechnological change, neo-Schumpeterian, Evolutionary and Structural perspectives haveasked “what type of innovation are we considering?” and “what are the structural, demand andinstitutional factors which help create or destroy jobs?”. Besides the quantity of employmentchange, quality aspects have received increasing attention, with the question on “what type ofjobs are created or destroyed by innovation?”. This line of research, often associated to anequilibrium view of the functioning of markets, has taken two main directions: “how does thecomposition of skills change” and “how does the wage structure change”, leading to a largeliterature on skill biased technical change and on wage polarisation.This chapter examines the evidence from empirical studies. First the perspectives, scope andtypes of innovations are considered, identifying the different employment effects they mayhave. Second, the effects on the quantity of employment are reviewed at the firm, industryand macroeconomic level. Third, changes in the quality of employment are examined,considering the effects on skills and wages, and the impact of organisational innovation, againat different levels of analysis. A summary of stylised facts concludes the chapter with adiscussion of future research and policy issues.1. Perspectives and scope of innovationsThe literature on innovation and employment has addressed different research questionsrooted in several streams of research. Table 22.1 summarises the main perspectives in termsof assumptions on the economic system, methodologies and levels of empirical analysis (for areview of theoretical approaches see Petit, 1995).The mainstream approach looks at innovation as a change in technology across the economy,leading to economic growth and employment. Early growth theory in a Keynesian perspective hadassumed that an (exogenous) new technology opens up investment opportunities, and that 1 For their comments and suggestions I wish to thank the editors, Charles Edquist, Bengt-AkeLundvall, Marco Vivarelli and several contributors to this volume. 1
  2. 2. investment increases output, income and employment.2 A new technology may have the effect ofsaving on capital (as well as on labour), leading to a lower capital-output ratio and to a fasterpotential rate of growth. The argument that new technology creates investment opportunities andincreases the marginal efficiency of capital, leading to steadily higher rates of growth, is parallel tothe view of Schumpeter and has returned fashionable in the 1990s in the debates on the ‘neweconomy’. The lack of an explicit conceptualisation of technology, either in the decisions of privateagents, or in the tools of public policy and its exogenous nature remain, however, a limitation ofKeynesian growth theories, including Solow’s growth model (see the chapter by Verspagen andLouca in this volume).Modern theories have approached the question of the employment impact of technology from twoperspectives. On the one hand, growth theories have moved from an exogenous view oftechnological change to efforts, in the new growth theory, to conceptualise innovation – proxied bytechnology, learning, and educational variables - as an engine of endogenous growth. Laboureconomists, on the other hand, have explained changes in employment (and wages) with jobdemography, macroeconomic factors, wage costs, bargaining modes and flexibility of labourmarkets, then moving to consider competitiveness and technology factors (see table 22.1).However, the fundamentally disequilibrating nature of technological change is usually treatedin a context which still assumes a general (or partial) equilibrium of markets, that is, alloutput finds its demand, and all workers ready to accept the current wage find employment.Technological change is often reduced to new production processes (and new productionfunctions) and models rarely envisage the emergence of product innovation. Whenemployment losses appear in such studies, they rarely lead to permanent (or structural)unemployment; they lead to downward adjustments in wages so that the additional jobless arereturned to work. If this cannot be found in the real world, then the responsibility is attachedto the lack of flexibility of labour markets, with excessive union power or institutionalrigidities such as the minimum wage.A more convincing approach to the study of innovation and its consequences is one thataddresses from the start the disequilibrium nature of economic change. This view has beendeveloped by neo-Schumpeterian perspectives, by Kaldorian, Structural and Evolutionaryapproaches (see table 22.1). Table 22.1 hereNeo-Schumpeterians have argued that advanced economies are witnessing the emergence of a newtechno-economic paradigm based on Information and Communication Technologies (ICTs) (see the2 Hansen (1964, ch.35) used the standard Harrod-Domar model - where g = s / c, the rate ofgrowth g is equal to the ratio of the share of savings s and the capital-output ratio c - to showthat a low propensity to save (i.e. a high multiplier) and a high capital-output ratio (i.e. a highaccelerator and large investment opportunities for a given ouput) increase the level of incomein the short term, with makets in a better position to reach full employment. A high rate ofgrowth, on the other hand, may require the opposite conditions. A high flow of savings and asmall capital-output ratio (i.e. a greater efficiency of capital, but lower investmentopportunities) make higher growth possible and the economy more efficient, assuming thatfull employment (and high income) is reached through an active public policy that fills thegaps in demand. Hansen wondered whether the higher rates of growth in Western Europecompared to the US in the 1950s and 1960s could be explained by the higher Europeansavings rate and the lower capital requirement (in countries characterised by high humancapital and extensive state intervention to stimulate demand in the economy). I thank RichardNelson for drawing attention to this work. 2
  3. 3. chapter by Fagerberg in this volume). A cluster of radical innovations in computers andtelecommunications has led to rapid improvements in technologies which have a widespreadapplication and diffusion potential and rapidly diminishing prices. Such a radical change in thenature and trajectories of innovations, and their widespread adoption is reshaping the economicstructure and international hierarchies, but is also affecting the way (and the place where) mosteconomic (and social) activities are carried out. Such powerful technological changes do create anddestroy a large amount of jobs. The question of how many (and where) are created and destroyeddepends from the highly dynamic process which shapes at the same time the content of specifictechnological innovations, the speed of adoption, the economic activities which are based on them,and the social relations which develop around them. Moreover, jobs lost and new jobs offered maytake place in different areas or require different skills, leading to mismatches. The speed ofadjustment is therefore crucial and makes the difference between frictional unemployment, easilyabsorbed by well functioning labour markets, and technological unemployment.Sustained and sustainable growth can be expected only once the mismatches between the newtechnologies and the old economic and social structures and institutions are overcome, with atwo-way adjustment. Innovation has to be adapted to social needs and economic demands;economic and social structures evolve under pressure from new technologies. Moreover, newtechnologies need to be matched by organizational changes, new institutions and rules,learning processes, emergence of new industries and markets, and expansion of new demand.Several studies on the emergence of technological paradigms and key technologies in the pasthave pointed out the long time required by the combination of all these elements before theirimpact on economic growth and employment could become evident (Freeman, Clark andSoete, 1982; Freeman and Soete, 1987; 1994). While extremely powerful in its explanation oflong term economic changes and historical evolution, this approach has yet to be“operationalised” with more specific questions on the type of innovation and on theinteraction with economic and employment variables.Box 1Technology and unemployment: a Classical debateSince the emergence of the industrial revolution, the extensive substitution of labour bymachinery incorporating the new technology of the time has led economists and policymakers to debate the economic and social consequences. At the end of the XVIII centuryJames Steuart drew attention to the difficulty of reabsorbing the unemployment caused bysudden mechanisation, in spite of the positive effects from the construction of new machinesand price reductions, and already envisioned a role for the government. Adam Smith linkedthe invention of machines to the division of labour and emphasised its labour saving effects.Jean Baptiste Say had less doubts on the ability of markets to adjust, while Thomas Malthusemphasised that the positive effects resulted from the strong demand dynamics experiencedby England at the time. The optimism of Classical economists in the early XIX centurycontrasted with the dramatic impoverishment of the English working classes – industrialworkers, small artisans and displaced peasants – who had started to organise trade unions andto launch Luddite struggles against the job losses and deskilling brought about bymechanisation. David Ricardo was convinced that the economy could compensate thenegative employment effects, but in a famous passage in the chapter “On machinery” added inthe third edition of his Principles of Political Economy and Taxation argued that “Theopinion, entertained by the labouring class, that the employment of machinery is frequentlydetrimental to their interests, is not founded on prejudice and error, but is conformable to thecorrect principles of political economy” (Ricardo, 1951:392). 3
  4. 4. The most articulated criticism of compensation theory was developed by Karl Marx whoemphasised the losses for workers in terms of jobs, skills, wages and control over their workresulting from the way mechanisation was proceding at the time. Arguing that unemploymentgrows as technical change displaces labour faster that the accumulation of capital demandsnew workers, Marx developed important insights on the functioning of capitalism. The driveto capital accumulation leads to a constant search for new production techniques and newproducts (a key starting point in Joseph Schumpeter’s theory of innovation). Highunemployment assures lower wages and greater control over workers, but capitalaccumulation ultimately encounters the problems of finding adequate markets and demand,and making adequate profits (for a reconstruction of the debate see Heertje, 1973; Vivarelli,1995).New products, processes and organisationsSchumpeter (1934) defined product innovation as “the introduction of a new good (...) or anew quality of a good” and process innovation as “the introduction of a new method ofproduction (...) or a new way of handling a commodity commercially”. He also analysed theemergence of new forms of organisation, the opening of new markets and of new sources ofmaterials.The development (or the adoption) of process innovations, introduced mainly through newinvestment, leads to greater efficiency of production, with savings in labour and/or capital,and with a potential for price reductions. The usual outcome is higher productivity and loss ofemployment; to the extent that process innovations increase product quality or reduce prices,a rise in demand (when elasticity is high) may result in more jobs.The introduction of product innovations can be based on internal innovative activities as wellas on the acquisition of new intermediate or capital goods. Products (or services) can beradical innovations (new to the world), incremental improvements on previous ones, orimitation of goods already produced in other countries or firms. Generally, productinnovations increase the quality and variety of goods and may open up new markets, leading(when elasticity is high) to greater production and employment. But new products can simplyreplace old ones, with limited economic effects, or be designed in order to reduce costs, withan impact similar to the one of process innovations (Pianta, 2001).New goods enter the economy as consumption goods, intermediate goods or investmentgoods, following the demands of consumers, firms and investors. Innovative investmentgoods have a dual nature; they start as new products in the industries producing them, butbecome process innovations in the industries acquiring them; their employment consequenceare likely to be positive in the machinery producing sectors, and negative in the industriesmaking new investments (Edquist, Hommen and McKelvey, 2001).The distinction between process and product innovations should not be brought too far. Mostinnovative firms introduce both at the same time (see Box 2 below), but in most firms andindustries (see the chapter by Malerba in this volume) it is possible to identify the dominantorientation of innovative efforts, associated to strategies of either price competitiveness (andmainly process innovations) or technological competitiveness (and mainly productinnovations).In addition to the technological changes in products and processes, organisational innovationis an equally important factor affecting the quantity and quality of employment, and is usuallyclosely linked to the introduction of new technologies (Caroli, 2001). Sections 2 and 3 belowwill consider the impact on job creation and loss of each of these types of innovation. 4
  5. 5. Innovation, imitation, adoption and useLooking at the emergence of new technologies from the perspective of the economic system,we have an innovation when a firm first markets a new product or introduces a new process;the road open to followers in the same industry (in other countries, too) is the imitation ofnew products (perhaps with incremental improvements, and adaptation to new users’ needs);firms in all sectors may decide on the adoption of new processes or use of new (intermediate)products generated in other industries (and/or countries) The latter two lead to the diffusion ofinnovations throughout the economy, in both production and consumption (see Hall in thisvolume).For empirical analyses, this distinction can be captured either by studies on the emergence ofparticular technologies, following their evolution through all the previous steps, or by studieson firms (or industries) based on surveys which identify product innovations novel to themarket (original innovators), those new to the firm only (imitators), and the introduction ofnew processes (adopters). Traditional indicators such as R&D and patents fail to capture alarge part of the latter two modes of technological development.An important strength of this perspective is that innovation is conceptualised from the start asa deliberate process of change sustained by firms’ efforts (learning, managing and spending)to develop knowledge, accumulate capital and access sources of innovation. In thisperspective, innovation is therefore a thoroughly endogenous process, highly specific to firmsand industries, affecting changes in both processes and products. This contrasts with thetraditional view of exogenous technological change in neoclassical growth models, with thereliance on knowledge spillovers and with the efforts of new growth theories to endogeniseinnovation through measures of R&D, learning or human capital. Such approaches tend toview innovation as the emergence of a new production technology changing capital/labourratios and productivity, concentrating therefore on process innovations, while the introductionof product innovations is rarely addressed in these models (see Verspagen and Louca in thisvolume).Box 2Evidence on innovation and employmentConceptual and empirical difficulties have long led economists to adopt a rather homogeneousview of innovation, described either by R&D expenditure (one of its inputs), or proxied bypatenting activity (one of its output). In the last decade the spread of innovation surveys inEurope and of surveys on panels of firms in the US has provided important new evidence onthe variety of innovative activities and on their economic and employment effect (see thechapter by Smith in this volume).Table 22.2 shows the key findings of the second Community Innovation Survey carried out inEurope (CIS 2, see European Commission-Eurostat, 2001), where large samples of firms(representative of the universe) have been asked whether they have introduced an innovation(either in products or in processes) in the period 1994-1996. The results are that 51 per cent ofmanufacturing firms and 40 per cent of service firms are innovative, with higher shares foundin Ireland, Denmark, Germany and Austria, and the lowest ones in Spain, Portugal, Belgiumand Finland. Considering manufacturing industry only, product innovations were introducedby 44 per cent of all European firms and process innovations by 39 per cent (the majority ofinnovative firms has introduced both); a stronger orientation towards product innovation isfound in Austria, Germany, Ireland, Sweden and the UK, while in Spain, Portugal and Italyprocess innovators are more frequent than product ones. 5
  6. 6. An important novelty of innovation surveys is the possibility to discriminate the relevance ofinnovations. Table 22.2 shows that 21 per cent of European firms have introduced productsthat are new to the firm’s market (not necessarily to the world market), showing that less thanhalf of all product innovations are important; the Netherlands, Denmark and Ireland were thecountries with above average shares. Finally, European firms reported that 12 per cent of theirturnover in 1996 was due to products new to the firms, a key indicator of the economic impactof innovations; Germany, Ireland, Sweden and Austria have the highest values.This selection of indicators offers an original perspective on the variety of innovativestrategies carried out by firms and on their market outcome. The key results are thatinnovation is present in services almost as much as in manufacturing; one fifth ofmanufacturing firms introduces products new to their market, and another fifth developsminor product improvements, while almost as many firms change their production processes;the impact of new products on sales, on the other hand is limited to 12 per cent. In allindicators important differences are found across countries, associated to the nature of theirindustrial structure and national innovation system (see the chapters by Smith and Edquist inthis volume). Table 22.2 hereThe latter variable offers the most accurate description of the economic relevance ofinnovations, and is related, in Figure 22.1, to the employment performance (average annualrate of change) of 20 manufacturing industries in four EU countries (France, Italy, theNetherlands and the UK). Looking at industry data for the share of new products in sales(drawn from the CIS 2-SIEPI database which provides data at the two digit industry level formajor countries), a very strong variability is evident, from more than 20 per cent in officecomputing and telecommunications, to close to zero in more traditional industries (again withimportant cross country differences due to national specialisations). Looking at overallemployment change, we first see that between 1994 and 2000 the majority of industries hashad job losses. The relationship between innovation and employment experienced in Europeanindustries looks like a positively oriented curve, but a closer look at the distribution of casesis important.ICT industries (computing, telecommunications, precision instruments and other transport,including aerospace) are generally in the top right quadrant, where new products in sales havea share higher than the average value for the EU (found in Table 22.1), and where no dramaticjob losses are found. As expected, industries characterised by the new technology (and by ahigh concentration of product innovations) show the highest impact of new products in theirturnover and better employment performances. In a few cases, however, moderate job lossesare found even in this group, as strong international competition may lead to the decline ofsome ICT industries in some countries.Traditional industries (textiles, wearing apparel, leather and a few others) tend to concentratein the bottom left quadrant, where a below average innovativeness (and a strong dominance ofprocess innovations) is matched by dramatic job losses.The remaining sectors, in the top left quadrant, combine a low or intermediate impact ofinnovation with modest job losses or substantial gains, showing again a positive association.While several factors alongside innovation affect employment change (macroecomicconditions, competitiveness, etc., see Pianta, 2000) the distribution of Figure 22.1 highlightson the one hand the generally positive link we can expect between innovation and jobs; on theother hand, it shows the presence of winners and losers in all industries, reflecing theimportance of national specialisations, of economic structures and the intensity ofinternational competition in open economies. Figure 22.1 here 6
  7. 7. 2. The effects on the quantity of employmentThe relationship between innovation and jobs is investigated in this section looking at theimpact on the quantity of employment, defined in terms of the number of existing jobs, or, inmore accurate analyses, in terms of the total hours of work. This link can be examined atdifferent levels of analysis: firms, industries and the aggregate economy. Table 22.3summarises the most relevant empirical evidence emerging from the literature.Direct effects at the firm levelFirms are the place where innovations are introduced and where they show their immediatedirect effects on employment. A growing literature has explored the issue with a variety ofmodels, national studies and panels of firms (for reviews, see Petit, 1995; Chennells and VanReenen, 1999; Spiezia and Vivarelli, 2002). Empirical work in this field has generally usedannual surveys of firms in panel data; however, panels are usually not representative of thewhole manufacturing industry, and in most cases leave out services altogether; therefore it isdifficult to generalise their conclusions.The evidence on the overall employment impact of innovation at the level of firms tends to bepositive; firms which innovate in products, but also in processes, grow faster and are morelikely to expand their employment than non-innovative ones, regardless of industry, size orother characteristics.3 The variety of innovative strategies and job creation and destructionpatterns has been highlighted in such studies, together with the firms’ characteristics(structural factors, flexibility, competences, etc.) which tend to be associated to betterperformances.A reverse relationship can also be considered. In the long run development of firms, phases ofrapid growth of employment may be seen as determinants of innovations as firms have tocope with the rigidity of production processes and increasing wages, while trying to capture,through greater productivity and quality, the opportunities of expanding markets.4Technological innovation is closely linked to organisational changes. A study on a large andrepresentative sample of French firms, has found that firms that have introduced both in theadoption of advanced manufacturing systems in the period 1988-93 have had a greateremployment growth than others, regardless of size or sector, and this positive effect was3 Generally positive effects of indicators of innovation on the number of jobs (once thecharacteristics of firms have been controlled for) have been found in the studies of UK firmsby Van Reenen (1997), who related a large panel of manufacturing firms to the SPRUdatabase of British innovations, and by Machin and Wadhwani (1991), Blanchflower,Millward and Oswald (1991), Blanchflower and Burgess (1999) using the British workplaceindustrial relations survey on the adoption of ICTs; the latter also found weaker similarevidence for Australia. For German firms, Entorf and Pohlmeir (1991) have relatedinnovation, export and employment in a cross section of firms, finding a positive effect ofproduct innovation and no effect of process innovations. Smolny (1998) has found similarresults with a panel of firms. Dutch firms investigated by Brouwer, Kleinknecht and Reijnen(1993) have shown an overall negative link between innovation and jobs, but where productinnovations were dominant better employment outcome have been found. In Norway, Kletteand Forre (1998) have found a negative association between R&D and employment at theplant level (including small units).4 This view is proposed by Antonelli (2001:173) in a study on the Italian car maker Fiat overthe time series 1900-1970, which shows that employment growth Granger-caused patentgrowth with different time lags. In turn, patent growth led to productivity growth. 7
  8. 8. greater than in firms that introduced organisational innovation only (Greenan, 2003, seesection 3 below).However, firm level studies on the innovation-employment link are unable to point outwhether the output and job gains of innovating firms are made at the expense of competitors,or whether there is a net effect on aggregate industry.5 It is often difficult to generalisebeyond the groups of firms investigated and to compare results across countries. When panelsare used, a large part of the jobs created or lost may be accounted by the entry or exit of firmsleft outside the panel. In order to address these issues we need to turn to industry levelstudies. Table 22.3 hereDirect (and a part of indirect) effects at the industry levelThe evolution of sectoral value added and employment may show whether the gains ofinnovative firms have been greater or smaller than the losses of non innovative ones. Ananalysis at the industry level therefore addresses not only the direct employment effects ofinnovation within firms, but also the part of the indirect effects which operate within theindustry. Such an impact includes first the competitive redistribution of output and jobs fromlow to high innovation intensive firms, and second, the evolution of demand (and thereforeoutput and jobs) resulting from the lower prices due to innovation, given the price elasticitiesof the industry’s goods.Studies on industries (see Table 22.3) have shown that the sources and opportunities for bothtechnological change and job creation are specific in individual manufacturing and serviceindustries, and such factors are key determinants of employment performances. The empiricalevidence shows that the employment impact is positive in industries (both in manufacturingand services) characterised by high demand growth and an orientation towards product (orservice) innovation, while process innovation leads to job losses. The overall effect ofinnovative efforts depends on the countries and periods considered, but in general is morepositive the higher is demand growth, the importance in the economy of highly innovativeindustries (both in manufacturing and services), and the orientation toward productinnovation. In open economies, countries with an economic structure of this type are likely toreceive a disproportionate part of the employment benefits of innovation; countries withstagnant economies and less innovative industries are likely to experience serious job lossesdue to technological change. Theories with roots in Schumpeter, Kaldor and Pasinettifocusing on structural change appear to explain effectively these empirical developments.The industry level may be the most satisfactory level of analysis as it is able on the one handto differentiate between the variety of technological regimes and firms’ strategies and, on theother hand, to bring in the demand dynamics of specific sectors, taking into account countrydifferences in economic structures.Demand factors are important because, differently from the case of firms, an industry’sdemand is constrained by the composition and dynamics of domestic and foreign demand.High demand growth leaves room for a variety of firms strategies and for better employment5 In France, Greenan and Guellec (2000) have found a positive relationship between bothproduct and process innovation and employment at the firm level, but at the industry levelonly the former played a positive role. Using data from the French household survey, Entorf,Gollac and Kramarz (1999) found that computer use reduces the risk of unemployment in theshort term, but not in the long term. 8
  9. 9. outcomes, while stagnant demand deepens the selection process among firms and emphasisesthe role of technological competition (Pianta, 2001).6The empirical investigations of the employment impact of innovation at the industry levelinclude studies using R&D or patenting as innovation proxies7, input-output models,8 andmore recent works based on innovation surveys (unfortunately, available only for Europe). Setin the context of the European debate on ‘jobless growth’ in the 1990s, the evidence points toan extensive process of restructuring in many manufacturing sectors where the growth ofvalue added is not matched anymore by increases in jobs (see Pianta, Evangelista and Perani,1996; Fagerberg, Guerrieri and Verspagen, 1999).Studies using innovation survey data matched to industry performances9 show that in Europeemployment change (in most cases a decrease) is affected by the dynamics of demand, ofstructural change, and the orientation of innovation towards new products, while a higherintensity of innovative expenditure per se has a negative impact on jobs. Once structuralchange and the nature of technological change are considered, a greater innovation (or R&Dand design) expenditure is associated to worse employment outcomes, suggesting a prevailingpattern of labour-replacing technological change.Similar findings come from a model where employment is affected by demand dynamics,labour costs and innovation variables associated with strategies of technological or pricecompetitiveness; in the context of the modest aggregate growth of the 1990s, Europeanindustries (in eight countries) were dominated by the latter strategy, with a prevalence ofprocess innovations, and generally negative effects on jobs (Antonucci and Pianta, 2002).The findings for service industries do not differ substantially from those of manufacturing.Studies on Italy have found an overall negative effect, concentrated among the largest firms,on low skilled workers, on capital intensive and finance-related sectors, and where the impactof ICTs has been most widespread. Smaller firms and technology-oriented activities show, onthe other hand, net employment gains (Evangelista, 2000; Evangelista and Savona, 2002,2003; see also the chapter by Miles in this volume).6 High demand is a necessary, but not sufficient condition for employment growth. In order toexpand jobs, demand and output have to grow faster than productivity, a condition which hasoften been found in the US, but less in Europe. Since the 1980s, the relation betweenproductivity growth and employment across industries has became negative (Appelbaum andSchettkat, 1995, Johnson, 1995).7 Typical of the former approach is the OECD Jobs Study (OECD, 1994) which investigated thehigh and persisting unemployment in advanced countries and downplayed the role of technologicalchange. The report emphasised the positive role of new technologies associated to structuralchange, and showed employment decline in several R&D intensive industries, and even greater joblosses in low-technology sectors.8 Meyer Kramer (1992) has used a model of the whole German economy to assess the impact ofICTs, proxied by direct and indirect R&D in 51 sectors; the findings suggest a generally negativeemployment impact, with some positive effects in higher technology industries.9 An early use of these data for Italian manufacturing industry is in Vivarelli, Evangelista and Pianta(1996) who have found a generally negative employment impact of technological change and theexpected contrasting consequences of product and process innovations. Pianta (2000, 2001) hasinvestigated the relationship between technological change and employment in five Europeancountries (Denmark, Germany, Italy, the Netherlands and Norway) in 1989-93 across 21manufacturing industries with variables accounting for changes in demand and industrial structure,for innovation intensity, and for the prevalence of product innovations. 9
  10. 10. The evidence at the sectoral level, especially for European countries, suggests a lessoptimistic view on the employment impact of innovation. The slow growth of the last decadehas represented a substantial constraint on the demand side, which has limited the potentialemployment benefits of technological change. While product innovation has confirmed itspositive effects on output and jobs, the increased international competition has pushed firmstowards restructuring, and process innovations have dominated several industries in manyEuropean countries, leading to a prevalence of labour saving effects.Direct and indirect compensation effects at the macroeconomic levelThe full view of the employment impact of innovation has to come from a macroeconomicperspective that can account and integrate all the indirect effects through which technologicalchange affects employment. This is the approach typical of the debate on “compensationmechanisms” which, since the times of Ricardo and Marx, has opposed those who argue thatthe economic system has in-built mechanisms which assure the recovery of jobs lost due toinnovation, and the critics who have pointed out the limitations of their effects and thepossibility of technological unemployment. It should be pointed out that most of this debateconceptualises technological change as the introduction of new capital goods, typical of XIXcentury mechanisation, therefore focusing on the impact that process innovations may have.Secondly, the equilibrium in product markets is generally assumed, i.e. no demand constraintis considered, following Say’s law. Recent research in this direction has relied on Kaldorianapproaches, on the work by Pasinetti (1981) and Boyer (1988a,b). A detailed treatment of thisissue is in Vivarelli (1995), who has summarized the compensation mechanisms and the waythey may (or may not) operate in the economy.The compensation mechanism via decrease in prices is one of the most important ones: newtechnologies may make lower prices possible, increasing international competitiveness andleading to greater output and to the recovery of the job losses due to the original innovation.This outcome, however, is contingent on the lack of demand constraints, on the decision offirms to transfer in lower prices the productivity gains due to the innovation, and on the lackof oligopolistic power in the relevant markets (Sylos Labini, 1969).The compensation mechanism via new machines may create new jobs in the industries wherethe new means of production are made, but the rationale for mechanisation is by definitionsaving on the use of labour, putting a limit to the relevance of this mechanism.The compensation mechanism via new investment argues that the temporary extra profitsavailable to the innovator may be turned into new investment; this however may expandproduction capacity and jobs, or may introduce additional labour saving effects.The compensation mechanism via decrease in wages is typical of the neoclassical view of thelabour market. As technological unemployment appears, wages would fall and firms wouldhire more workers. This mechanism however is based on strong assumptions on the feasibilityof any combination of labour and capital, competitive markets, flexibility of wages and labourmarkets.The compensation mechanism via increase in incomes operates in the opposite way, throughthe demand effects of the distribution of part of the gains from innovation to higher wages, asit has happened in large oligopolistic firms in mass production industries. However, wageincreases can hardly be large enough to sustain additional aggregate demand.Finally new products may lead – as discussed above - to new economic activities and newmarkets (welfare effects) or, on the other hand, they may simply replace existing goods(substitution effect, see Katsoulacos, 1986).Aggregate empirical patterns for the US are examined in a descriptive way by Baumol andWolff (1998). Considering five innovation indicators for the whole economy and their link tothe structure and changes in unemployment in the US in the 1950-1995 period, they conclude 10
  11. 11. that faster innovation leads to a higher ‘natural rate of unemployment’ and to longer frictionalunemployment.10Using aggregate models of simultaneous equations, a number of empirical test on therelevance of compensation mechanisms have been carried out. In a general equilibriumperspective, Layard and Nickell (1985), Nickell and Kong (1989) and Layard, Nickell andJackman (1991; 1994) have found that the working of various compensation mechanismsruled out the possibility of technological unemployment in the UK.Building on Boyer (1988a,b) and on the Regulationist approach, Vivarelli (1995) hasdeveloped a model for testing the compensation mechanisms in the US and Italy. Consideringtotal work hours in the economic system as the employment variable, he has found that themechanism via decrease in prices is the most effective one, and that the positive effects ofnew products and labour markets operate in the US (where new jobs are created), but not inItaly, where net job losses have been found. This approach has been further developed bySimonetti, Taylor and Vivarelli (2000), who have considered four countries, and by Simonettiand Tancioni (2002) who have developed a model for an open economy in the case of the UKand Italy. All have found a differentiated impact of compensation mechanisms.While this approach is the most comprehensive and satisfactory for explaining the overallimpact of technological change on employment, the complexity of the construction of themodel, the problems in specifying all relevant relationships and the lack of adequate data limitthe feasibility of this approach. The evidence available points to a differentiated impact ofinnovation depending on countries’ macroeconomic conditions and institutional factors; theemployment impact generally is more positive the higher is the ability to generate newproducts and to invest in new economic activities, and the stronger is the effect of pricereduction, leading to increased demand.Simulation studiesThe employment impact of innovation has also been studied with a simulation approach.Leontief and Duchin (1986) have estimated that the diffusion of computer technology andautomation in the US economy would have negative employment effects; they used an input-output model where, however, high assumptions on productivity growth were made and nodemand dynamics was allowed.11A study on the impact of microelectronics in forty sectors of the UK economy in the 1980sshowed the possibility of both net job gains and losses, depending on the assumptions on thespeed of diffusion (Whitley and Wilson, 1982). For Germany, Matzner and Wagner (1990)carried out a complex study of the effects of technological change in all fields, finding amoderately negative impact on jobs, but a positive impact on international competitivenesswhich in turn protects employment.A different approach - a general equilibrium model with a sectoral structure, which assumesfull employment - has been used for simulating the employment impact of different scenariosof technology-based productivity growth and of composition of consumption, in a recentstudy by IPTS-ESTO (2002) on the European Union. The results show an overall positiveimpact on jobs, differentiated according to the alternative sectoral distributions of R&D and10 They argue that “The evidence supports the conclusion that an increase in the pace of innovation(all else equal) will raise both the natural rate of unemployment and the average length of timeduring which an unemployed worker is ‘between jobs’” (p.10).11 Their assumptions on productivity growth were based on the improvements in theengineering performance of robots; the long time lags in their adoption and their poor initialperformance have meant that their actual economic effect on productivity has been muchlower. 11
  12. 12. innovation efforts; the best outcomes result from the concentration of efforts in hightechnology industries.While they are interesting as explorations of alternative futures, the weaknesses of suchsimulations are in their reliance on models which severely constrain from the start thepossibility of identifying either technological unemployment (when general equilibrium isused) or most compensation effects (when input-output models are used), and on thearbitrariness of the assumptions made on the diffusion and productivity of new technologies.Summarising the results of this section, both sectoral and aggregate studies generally pointout the possibility of technological unemployment, which emerges when industries orcountries see the prevalence of process innovations in contexts of weak demand. Firmsinnovating in both products and processes may be successful in expanding output and jobsregardless of the economic context, but often do so at the expense of non innovating firms.The specificities of industries, countries and macroeconomic conditions are crucialdeterminants of the results obtained in empirical studies.All the analyses of this section refer to national economies, sometimes even assuming aclosed economy. When we consider an open economy, the picture becomes more complex, ason the one hand innovation may lead to competitiveness and exports, thus weakening thedemand constraint; on the other hand domestic demand may increase imports when foreigncompetitors are more innovative in terms of price or quality. Unfortunately, no empiricalanalysis has so far addressed the innovation-employment question in a truly global dimension;as several manufacturing sectors are highly internationalised, the introduction and diffusion ofinnovation leads to job creation and losses in a large number of different countries, with acomplex distibution of the benefits and costs of technological change. The case of developingcountries in this context is particularly interesting, as new technologies are at the centre of thestructural changes and dynamic learning economies typical of catching up countries, whichhowever face greater difficulties in capturing the employment benefits of technologicalchange (see Karaomerlioglu and Ansal, 2000 and the chapter by Fagerberg and Godinho inthis volume).3. The effects on the quality of employmentMost approaches that assume equilibrium in labour markets (and therefore no technologicalunemployment, see Table 22.1) have disregarded the effects of innovation on the quantity ofjobs, and have focused on the relative composition between skilled and unskilled workers andon wage polarisation. Moving away from an undifferentiated view of employment is awelcome development, although the conceptualisation and the empirical definitions in thisfield are often crude, based either on educational levels or on blue/white collar occupations,or on both. The importance of learning, competences and skills has also been emphasised bythe evolutionary approaches to innovation, although few empirical investigations have beencarried out.A large (mainly) US literature on skill biased technical change (reviewed in Acemoglu,2002:7) argues that technical change is biased towards skilled workers as it replaces unskilledlabour and increases wage inequality. In fact, a strong complementarity between technologyand skills has characterised most of the XX century, when innovation has probably alwaysbeen skill biased, in contrast to the unskill bias typical of the XIX century, whenmechanisation had led to the deskilling of artisans (see Braverman, 1974).1212 Petit (1995) suggests that techniques and human labour had a complementary relationship in pre-industrial times, following strong social norms; industrialisation has been characterised by the 12
  13. 13. This section considers the streams of research that have found evidence on thecomplementarity between ICTs and high skills, but also on the deskilling process associatedto the greater control over production made possible by ICTs. The effects on the skillstructure of employment usually lead to changes in the wage structure where a tendency topolarisation is generally found. Increasing attention is also devoted to the impact thatorganisational change – alone or combined with technological change - has on skills, wagesand employment (for reviews see Chennells and Van Reenen, 1999; Sanders and ter Weel,2000). Table 22.4 summarises the most relevant literature. Table 22.4 hereSkill biased technical changeThe relative importance of skilled work has increased in the past century, rooted in thechanges in economic structures, technologies and in the growing availability of educatedworkers. Many studies – mainly on US firms and industries - argue that in the last twodecades this process has accelerated as a result of the introduction of information technologyand computers. The issue has generally been investigated using a factor substitutionframework, showing that direct or indirect measures of technology are important explanatoryfactors for the relative increase of skilled labour (see Berman, Bound and Griliches, 1994;Autor, Katz and Krueger, 1998). One stream of work has compared the effects of technologywith those of increased international trade, finding that the former accounted for most of thefall in demand for less skilled workers (see Berman, Bound and Machin, 1998). Other studieshave found that new technologies are adopted more in plants with more skilled workers, butdo not increase the demand for skills (Doms, Dunne and Trotske, 1997).The dominant findings of the econometric literature on skill bias in industries, firms andindividuals, using direct measures of technological change, is that the diffusion oftechnologies has a strong skill bias effect, while it has a less evident effect on wages (see thereview in Chennells and Van Reenen, 1999).However, little research has addressed the effects of different directions of technologicalchange and types of innovation (e.g. product versus process) as Sanders and ter Weel(2000:34) point out. Evidence has been found on upskilling within the firms and the industriesinvestigated, but also between firms and sectors, as the most dynamic ICT-related industriesaccount in all countries for large (relative) increases of skilled employment.In Europe such structural factors may be particularly strong. Machin (1996) for the UK andPiva and Vivarelli (2002) for Italy find some upskilling over the past decades, although in thelatter case organisational innovation appears to be the most important factor, whiletechnology and foreign direct investment play a secondary role. Indirect evidence on thelimited pace of ICT diffusion and upskilling in European economies, and on the importance oforganisational changes and work intensification has come from the third Survey on EuropeanWorking Conditions (European Foundation for the improvement of living and workingconditions, 2001).13substitution of machines for labour on the basis of profit seeking; post-war rapid growth has beenassociated to technologies designed to overcome labour shortages and increase productivity, whilethe current emergence of a new technological system based on ICTs is marked by uncertainties.13 The survey has interviewed 21,500 workers in all EU countries and has found that thenumber of people working with computers (at least one quarter of the time) has increased onlymarginally from 39 per cent in 1995 to 41 per cent in 2000, while those using computers allthe time are 19 per cent. Little change has taken place over that period also in the workers’perception of their skills: 8 per cent regard the demands of the job as too high for their skills 13
  14. 14. However, when more refined measures of skill are used, the evidence is more controversial.Moving from the simple measures of blue or white collar jobs, or of years of schooling, tomore refined indicators of skills, including cognitive (typical of technical staff), interactive(typical of supervisory staff) and motor competences (typical of manual workers) in USindustries between 1970 and 1985, Howell and Wolff (1992) have found that expenditures oncomputers and new investment were associated to raising demand for high cognitive skillworkers, although with differences across occupations and industries. Using these indicatorslittle support is found for the hypothesis of technology-skill complementarity.The aggregate evidence for the US economy is also weak. Howell (1996) rejects the idea of alink between computerisation, upskilling and wage inequality and finds that major shifts inskill structure happened between 1973 and 1983, and little change took place afterwards,when diffusion of ICTs accelerated and computer-relate investment per employee increaseddramatically.14 In recent years the shares of high skilled blue collars and low skilled whitecollars have declined faster, and in services little change in the skill structure took place. Theconsequences of ICT investment since the mid 1980s therefore appear to have affected thelow skilled white collars (mainly female) much more than the lowest blue collar skills.Wage polarisation and the labour marketThe effects of technological change on skills and wages are often treated together; morespecific studies have investigated the evolution of wage differentials in three contexts: acrossindustries with varying technological and capital intensities; among workers and firms usingor not using computers or new technologies; among workers or social groups with differenteducational levels (see Sanders and ter Weel, 2000; Acemoglu, 2002). Evidence ofpolarisation is generally found, although the relationship to technological change isconceptually and statistically less robust, as computers are more likely to be used by morecompetent workers who already earn higher wages (Chennells and Van Reenen, 1999). On theother hand, the technology-wage polarisation link has been questioned by studies pointing outthe lack of an acceleration in these effects in recent years, the importance of sectoral shiftsand the effects of growing international trade on wages (Mishel and Bernstein, 1996; Addisonand Teixeira, 2001).and 11 per cent as too low. However, work intensity has increased, as the share of workersreporting working at very high speed during at least one quarter of their time has increasedfrom 48 per cent in 1990 to 56 per cent in 2000 (this is closely correlated to health problemsand injuries at work). The share of workers which have control over their pace and methods ofwork has remained high and stable at about 70 per cent between 1995 and 2000, while only 44per cent (including self-employed) have control over their working time (EuropeanFoundation for the improvement of living and working conditions, 2001).14 Computer-relate investment per employee rose from 150 US$ in 1982 to about 1000 US$ in 1992for both manufacturing and services (id:292). The share of low skilled blue collar US workers inmanufacturing declined from 45.1 per cent in 1978 to 39.7 per cent in 1982 and has then remainedstable ending in 1990 with a 41 per cent share. In the late 1980s, when ICTs became important, themost serious reduction in the employment shares concerned the high skilled blue collars (who since1978 had a stable share around 21.7 per cent and declined after 1985 to 19.7 per cent in 1990) andthe low skilled white collars (who stayed stable around 12 per cent until 1985 and then declined to10.6 per cent in 1990). At the opposite end, the share of high skilled w collars had an increase in hitethe 1978-1983 period (from 19.5 to 24.1 per cent), then was stable until a raise in 1989-1990 to 25.8per cent (id:299, tables 1 and 2). 14
  15. 15. In comparing US and European results a paradox emerges. Empirical patterns show that in thelast two decades low skill and low wage jobs have increased much more in the US than inEurope, associated to a faster growth of the supply of labour, resulting from demographictrends, and to a greater polarisation of the wage structure. An alternative explanation of thesepatterns may be proposed, arguing that skill biased technical change is more a European thanan American story. With a stagnant labour supply and more educated workers, with sloweraggregate growth in the past two decades and greater competitive pressure, most Europeanfirms and industries appear to have taken the road of unskilled labour saving innovations.This has raised wage inequality, although the strong European labour market institutions havelimited polarisation in comparison to the US.In contrast, The Unites States has experienced faster growth of population, labour supply andGDP, with the expansion of new sectors based on product and service innovations, in morecompetitive labour markets where less regulation on minimum wages and union power arefound. This has resulted in a faster growth of new jobs (compared to Europe) at the top andbottom end of the skill structure, and this polarisation has been amplified in terms of wageinequalities by the lower regulation of US labour markets.15The effects of organisational innovationStudies on organisational innovation associated to the introduction of new technologies havelong identified opposing patterns of change. On the one hand there is evidence of deskilling ofworkers as machines incorporate human knowledge and make it possible to use cheaper andless qualified labour, a pattern that has emerged since the origins of modern industry and nowfrequently observed in manufacturing industries and low skill services (Braverman, 1974). Onthe other hand there is evidence of an upskilling of some occupations, resulting from thecomplementarities between new ICT technologies and skills (Bresnahan et al., 2002; see alsothe chapter by Lam in this volume).Studies on several countries collected in Adler (1992) find that both processes take place as aresult of different strategies of firms, suggesting that “the use of new technologies will ingeneral be more profitable when entrusted in to more highly skilled employees” (id:3) withbroader roles, greater competences and continued learning. However, it has been argued that“there is a fundamental contradiction between the potential of computerization to enrichworking life and increase productivity and the development of the technology in the pursuit ofauthoritarian social goals” (Shaiken, 1984:5) as management has often introduced newtechnologies and shaped work organisation with the primary aim to increase control overworkers (see also Noble, 1983, 1984).In an extensive survey of organisational change in US manufacturing and service firms,Appelbaum and Batt (1994) find a large diffusion of a variety of new management practices,often associated to the introduction of new technologies, resulting in improved productivity,quality and cost savings for firms, but no information is provided on the impact onemployment levels. Similar findings have emerged from a study on a large number of USestablishments, mainly in manufacturing, surveyed in 1993 and 1996, with both crosssectional and longitudinal analyses. Both technology adoption (such as the number of non15 This may amount to a shift in “wage norms”: “in the face of mounting competition, employersreduced unit labour costs and increased flexibility in the production process by following the ‘lowroad’ – lower wages, little training, and fewer permanent employees” (Howell, 1996:301).Combined with a large use of part time and temporary workers, anti-union practices, relocation tolow wage sites and inflows of low wage foreign workers, these developments have led to anincrease in the supply of labour competing for low skill jobs, leading to a major fall in their wages,accompanied by a 25 per cent reduction in the real minimun wage over the 1980s (id:292). 15
  16. 16. managers using computers) and new workplace practices are closely associated to higherproductivity and wages (Black and Lynch, 2000).Several European studies (Caroli and Van Reenen, 2001 on France and Britain; Greenan, 2003on France; Piva and Vivarelli, 2002 on Italy) have shown that organisational innovation ismore important than technological innovation in shaping changes in occupational structureand skills. This is generally not associated to an increase in the number of employees, withthe exception of management occupations.16Organisational and technological changes in services, on the other hand, have reflected theopportunities offered by ICTs in order to overcome time and space constraints in theprovision of services, leading to major flows of job creation and destruction, and to rapidlychanging skill requirements. A variety of strategies of restructuring, emergence of networks,subcontracting and outsourcing, has resulted, leading to polarisation effects in terms of skillsand wages (Petit and Soete, 2001b; Frey, 1997).The rather fragmented evidence so far available on organisational innovation suggests that itplays a crucial role alongside technological innovation in shaping employment outcomes. Thetwo can have a complementary relationship (especially when a virtuous circle of growth is inplace), or represent alternative paths for adjustment (especially when restructuring and joblosses are under way).A broader viewWhile little doubt exists on the contribution of technological change to the long termimprovement of the skills and wages of the more fortunate workers, the specific effects foundin particular countries and periods are the result of economic structures, of the prevailingstrategies of firms, of the operation of labour markets rules and institutions and of nationalpolicies. The link between innovation and employment has to be set in this broader context,bringing back a macroeconomic dimension which is missing in current literature. Littleresearch has moved beyond the question of changing relative wages and addressed thebroader distribution effects of technological change between profits, rents, wages, shorterwork hours and lower prices.So far attention has been devoted to the demand for labour only. New technologies interactwith broader social changes leading to effects also on the supply of labour. Key processesaffecting the quantity of supply include the slow demographic growth, immigration flows, anageing population and the forms of women’s presence in the labour force; the quality oflabour supply results, on the other hand, from education and training, learning processes andaccumulation of competences.A major process of change has recently invested the forms of employment, with firmsstrategies and governments policies leading to the decline of the traditional model of fulltime, life time, waged (and unionised) employment and to the rapid growth of flexible,temporary, part time, subcontracted work.On the other hand, social dynamics has led to demands for reducing the ‘precarisation’ ofwork, maintaining welfare protection, for shorter working time in Europe, greater training andlife long learning, more meaningful jobs and to the development of socially useful activities16 A survey on Italian firms in the Bergamo district has found a limited diffusion of new workorganisations and models of Human Relation Management, due to a general lack of skilled labourand to a failure by firms to use the potential of human resources in learning, compentence buildingand problem solving (Leoni et al., 2003). A survey of firms in the Reggio Emilia district hasinvestigated the determinants of both technological and organisational innovation, finding that theyare more likely to develop in firms having a higher profits, high shares of blue collar workers, lowhierarchies and greater management-worker interactions (Antonioli et al., 2003). 16
  17. 17. carried out in the ‘third sector’ of non profit organizations. The diffusion of new technologiesand the combination of technological, organisational, institutional and social innovations willhave to provide answers also to these broader social demands.4. A summary of stylised facts and research directionsThis chapter has examined the relationship between innovation and employment and revieweda large empirical literature addressing a variety of research questions at different levels ofanalysis. The complexity of the issues is such that no single approach can account for thedirect and indirect consequences of technological change, for its effects on the quantity andquality of labour employed in the economy. Theories and empirical research have to proceedin parallel, with a close interaction between concepts and measures, hypotheses and tests,building links to related research areas. In this final section, a few stylised facts emergingfrom the empirical evidence are summarised, leading to possible directions for future researchand policy.A few stylised factsThe never ending race of innovation and employment. The evolution of (most) economies showsthat when growth, structural change and demand dynamics take place together, in the long run thejobs lost to technological change are found elsewhere in the economy. If no innovation took place,economic activities facing competitive pressure would cut costs, wages and eventually jobs. Thekey question is how long does it take, and how the pace of innovation and diffusion eliminating oldjobs compares to the pace of development of new economic activities creating new jobs. Thismakes the difference between the frictional unemployment of neoclassical economics and thetechnological unemployment of disequilibrium theories.Technological unemployment can happen. On the basis of the available evidence currenttechnological change may be a cause of unemployment. There is no automatic mechanismensuring that a national economy is able to fully compensate for innovation-related job losses.This applies in particular to Europe in the 1990s.The type of innovation is important. The evidence shows that it is essential to discriminatebetween product innovation (novel or imitative) which has a generally positive employmentimpact, and process innovation (adoption and use of new technologies) usually with negativeeffects; these findings emerge regardless of the theoretical approaches used.Organisational innovation is closely linked to technological change. Changes in organisationsreflect a variety of business strategies ranging from internal or external growth torestructuring and downsizing (leading to the job losses frequently documented by empiricalstudies); organisational innovation can play a complementary role to technological change, orcan represent an alternative strategy in firms’ restructuring.The skill bias effect is relevant. Unskilled jobs have long been declining in absolute terms (inEurope) or growing slowly (in the US), while skilled jobs for educated workers are created ata faster pace in most countries, and are associated to greater innovative efforts. Thediscussion remains open on the specific effect of ICT-related innovation since the 1980s andon the role of the supply of more educated labour. 17
  18. 18. Wage polarisation is strong. Since the 1980s most countries have experienced a growingdivide within wages (and more generally in incomes). This is the result of changes ineconomic structures, in firms’ strategies and in government policies. Technological change iscertainly part of the story, but the specific link between innovation and wage differentials hasbeen questioned; a broader view of the evolution of labour markets, employment forms, socialrelations and national policies may explain in a more convincing way growing wage andincome inequalities.Aggregate demand and macroeconomic conditions are important. They play a key role increating the conditions for a positive impact of technological change on employment, incomeand distribution. While the role of demand has generally been downplayed in the innovationliterature, it may help explain the negative impact of innovation found in Europe in the lowgrowth decade of the 1990s.Innovation interacts with trade. In open economies trade is an important factor – alongsideinnovation – affecting employment and wages. The evidence available (mainly for the US)shows that job losses (especially for unskilled workers) and lower wages are explained bytechnological change more convincingly than by increased trade. But the interactions betweeninnovation and trade shape competitiveness, the direction of technological change, theevolving division of labour and the employment outcomes.The national innovation system is the source of change. A country’s technologicalopportunities and ability to develop long term learning and innovating capabilities are rootedin the nature and characteristics of its national innovation system. Its strengths, orientationand priorities are likely to be reflected in the employment effects of a country’s innovativeefforts.Labour market conditions and institutions matter. The employment outcomes of technologicalchange depend on the way job creation and destruction takes place, wages are set, learning,flexibility and welfare protection are managed. On the other hand, labour market institutionsinfluence the supply of labour, which should match the skill and competence requirementsemerging with new technologies.Directions for future researchThe following research issues have emerged from this chapter as promising directions forfuture studies.In terms of method, both theories and empirical studies should make efforts to developexplanations and predictions which are consistent at the levels of analysis of firms, industriesand aggregate economy. Too often the findings at one level of analysis contradict the patternsat the more aggregate level. Ideally, this would require firm level data to be representative ofwhole sectors, and industry coverage to account for all the economy (i.e. including services).Innovation survey data, when accessible, could make this effort possible.Studies on firms could address the innovation-employment question in the context of theevolution of firms and market structures through the processes of diversity generation(through innovation) and selection (in the marketplace, Nelson and Winter, 1982), as thepotential for job creation depends of the dynamics of firms’ growth and new firms generation. 18
  19. 19. An important challenge to innovation studies is to extend to services studies generallyconfined to manufacturing industry. From a theoretical perspective, an analysis linkingstructural and technological (and organisational) change is highly appropriate to account forthe expansion of services in advanced economies. In empirical terms, the increasingavailability of industry level data also for services (such as in European innovation surveys)makes this extension possible. As most new jobs are created in services, this is highly relevantalso from a policy perspective.The need to devote greater attention to organisational innovations and to link them totechnological ones has already been pointed out; this appears as a major direction for futurestudies.Bringing the analysis of the innovation-employment link into the macroeconomic context ofan open economy is also important, in order to highlight the interactions betweentechnological developments on the supply side and growth potential on the demand side, aswell as the interactions between innovation and trade in shaping growth and employmentoutcomes.As international production, especially in manufacturing industries, becomes a key aspect ofindustrial structures, an additional level of analysis, looking at individual industries at theglobal level, may emerge. Data constraints are serious, but substantial evidence is nowavailable on the international production networks of multinational corporations and the linkto employment variables across the relevant countries may become possible.The appropriate labour market arrangements for favouring a virtuous circle betweeninnovation and a high quantity and quality of jobs and wages, through greater learning,competence building and improved working conditions is an additional area for futureresearch.Finally, as already suggested, research could address the distribution effects innovationthroughout the economy, which are much deeper than their effects on wage polarisation andhave a major policy relevance.Policy considerationsFive key principles for policy emerge from the evidence gathered in this chapter (see Vivarelli andPianta, 2000 and Lundvall and Borras in this volume).The first is the need for a pro-employment policy, focusing on employment friendly innovation; thedistinction between product and process innovations should inform policy in this field. Supply-sideincentives and funds for innovation should introduce a clear focus on the type of innovativeactivities more likely to result in market-creating new products, rather than in labour-displacing newprocesses.The second one is the need for targeting the industries with the greater potential for growth andemployment, and for specific actions directed at the needs of individual industries, with policiesrelying less on the traditional supply push incentives, and stimulating more the demand pull effectsof new potential markets and the active role of users.The third one is the need to expand education and learning throughout the economy - in schools,universities, in continuing education and on the job - in order to accelerate social change andsupport the demand for higher skills coming from innovative economies, industries and firms. 19
  20. 20. The fourth one is the need for taking seriously the systemic nature of innovation and the role ofnational innovation systems (see the chapters by Edquist and by Lundvall and Borras in this book).This implies a strong coherence between industrial, technology, learning and macroeconomicpolicies.The fifth one is the need for policies on the distribution of the productivity gains madepossible by new technologies. So far innovation has mainly benefited firms and consumers, inthe form of higher profits and lower prices. The result has been an increasingly unevendistribution of wages and incomes. If we want all to share in the benefits promised by the newtechnologies, it is vital that these negative trends be reversed through the pursuit of a newgeneration of policies. BibliographyAcemoglu, D. (2002), Technical change, inequality and the labor market, Journal of EconomicLiterature, 40,1, 7-72Addison, J. and Teixeira, P. (2001) Technology, employment and wages. Labour, 15, 2:191-219.Adler, P. (1992) (ed.) Technology and the future of work, New York, Oxford University Press.Antonelli, C. (2001) The microeconomics of technological systems, Oxford, OxfordUniversity Press.Antonioli, D., Mazzanti, M., Pini, P. and Tortia, E. (2003) Organisational and technologicalinnovations in manufacturing firms: diffusion and determinants. Quaderni del dipartimento dieconomia, istituzioni, territorio, University of Ferrara.Antonucci, T. and Pianta M. (2002) The employment effects of product and processinnovations in Europe. International Review of Applied Economics, 16, 3, 295-308Appelbaum, E. and Batt, R. (1994), The new American workplace, Ithaca, ILR Press.Appelbaum, E. and Schettkat, R. (1995) Employment and productivity in industrializedeconomies, International Labour Review, 134, 4-5: 605-623.Archibugi, D., Evangelista, R. and Simonetti, R. (1994) On the definition and measurement ofproduct and process innovations, in Y. Shionoya and M. Perlman (eds) Innovation inTechnology, Industries and Institutions. Studies in Schumpeterian Perspectives, Ann Arbor,The University of Michigan Press, 7-24.Autor, D., Katz, L. and Krueger, A. (1998) Computing inequality: have computers changedthe labor market? Quarterly Journal of Economics, 113, pp.1169-1214.Bartel, A.P. and Lichtenberg, F.R. (1987). The comparative advantage of educated workers inimplementing new technology, Review of Economics and Statistics, vol. 69, 1-11.Bartel, A.P. and Lichtenberg, F.R. (1991) The age of technology and its impact on employeewages. Economics of Innovation and New Technology, 1, 215-31Bartel, A. and Sicherman, N. (1999) Technological change and wages: an interindustryanalysis. Journal of Political Economy, 107:285-325.Baumol W. and Wolff, E. (1998), Side effects of progress. How technological change increases theduration of unemployment. Jerome Levy Economics Institure of Bard College, Public policy brief 41.Berman E., J. Bound, Z. Griliches, (1994), Changes in the Demand for Skilled Labor WithinUS Manufacturing Industries: Evidence from the Annual Survey of Manufactures, QuarterlyJournal of Economics 109:367-398.Berman, E., Bound, J. and Machin, S. (1998) Implications of skill biased technologicalchange: international evidence. Quarterly Journal of Economics, 113:1245-79.Black, S. and Lynch, L. (2000) Whats driving the new economy: the benefits of workplaceinnovation. NBER working paper 7479.Blanchflower, D., Millward, N. and Oswald, A. (1991) Unionisation and employmentbehaviour, Economic Journal, 101, 815-834. 20
  21. 21. Blanchflower, D. and Burgess, S. (1999) New technology and jobs: comparative evidencefrom a two country study, Economics of Innovation and New Technology.Boyer, R. (1988a). Technical Change and the Theory of Régulation, in Dosi et al., 67-94.Boyer, R. (1988b). Formalizing Growth Regimes, in Dosi et al., 608-35Braverman, H. 1974. Labour and Monopoly Capital, New York, Monthly Review Press.Bresnahan, T.F. (1999) Computerisation and wage dispersion: an analytical reinterpretation.The Economic Journal, 109:F390-F415.Bresnahan, T.F., Brynjolfsson, E. and Hitt, L.M. (2002) Information technology, workplaceorganization and the demand for skilled labor: firm-level evidence, Quarterly Journal ofEconomics, 117.Brouwer, E., Kleinknecht, A. and Reijnen, J.O.N. (1993). Employment Growth andInnovation at the Firm Level: An Empirical Study, Journal of Evolutionary Economics, vol. 3,153-59.Caroli E., (2001), New Technologies, Organizational Change and the Skill Bias: what do weknow?, in P. Petit and L. Soete (eds), 259-292.Caroli, E. and Van Reenen, J. (2001) Skill biased organizational change? Evidence from apanel of British and French establishments. Quarterly Journal of Economics, 116, 4:1149-1192.Casavola, P., Gavosto, A. and Sestito, P. (1996) Technical progress and wage dispersion inItaly: evidence from firms’ data. Annales d’Economie et de Statistique, January-June, 387-412Chennells, L. and Van Reenen, J. (1999), Has technology hurt less skilled workers? An econometricsurvey of the effects of technical change on the structure of pay and jobs. Institute for Fiscal Studiesworking paper 27, London.Cyert R.M. and Mowery, D.C. (eds) (1988) The impact of technological change onemployment and economic growth. Cambridge, Mass, Ballinger.Doms, M., Dunne, T. and Trotske, K. (1997), Workers, Wages, and Technology, QuarterlyJournal of Economics, vol. 112, pp. 253-89.Dosi, G., Freeman, C., Nelson, R., Silverberg G. and Soete L. (eds) (1988) Technical Changeand Economic Theory, London, Pinter.Edquist, C., Hommen, L. and McKelvey, M. (2001), Innovation and employment: productversus process innovation, Cheltenham, Elgar.Entorf H., Gollac, M. and Kramarz, F. (1999) New technologies, wages and worker selection,Journal of Labor Economics, 17, 3 464-491.Entorf, H. and Kramarz, F. (1998), The impact of new technology on wages and skills:lessons from matching data on employees and their firms. Economics of Innovation and NewTechnology, 5 (2-4), 165-197.Entorf H. and Pohlmeir, W. (1990) Employment, innovation and export activity, in Florens, J.et al. (eds) Microeconometrics: surveys and applications. Oxford, Basil Blackwell.European Commission-Eurostat (2001) Statistics on innovation in Europe. Data 1996-1997.Luxembourg, European Commission.European Foundation for the improvement of living and working conditions (2001) ThirdEuropean survey on working conditions 2000. Dublin, European Foundation for theimprovement of living and working conditions.Evangelista R. (2000) Innovation and employment in services: results from the Italianinnovation survey, in Vivarelli and Pianta (eds)Evangelista, R. and Savona M. (2002), The impact of innovation on employment and skill inservices. Evidence from Italy. International Review of Applied Economics, 3, 2002.Evangelista, R. and Savona, M. (2003), Innovation, employment and skills in services. Firmand sectoral evidence. Structural Change and Economic Dynamics (forthcoming).Fagerberg, J, Guerrieri, P and Verspagen, B (eds) (1999) The Economic Challenge forEurope. Adapting to innovation based growth. Northampton, Elgar 21
  22. 22. Freeman, C. and Louca, F. (2001) As time goes by. From the industrial revolution to theinformation revolution. Oxford, Oxford University Press.Freeman, C. and Perez, C. (1988) Structural crises of adjustment, business cycles andinvestment behaviour, in Dosi et al.Freeman, C. and Soete, L. (1994) Work for all or mass unemployment?, London: PinterFreeman, C. and Soete, L. (eds) 1987. Technical Change and Full Employment, Oxford, BasilBlackwellFreeman, C., Clark, J. and Soete, L. (1982) Unemployment and Technical Innovation, London, PinterFrey, L. (1997) Il lavoro nei servizi verso il secolo XXI, Quaderni di Economia del Lavoro, 57, Milan,Angeli.Greenan, N. (2003) Organisational change, technology, employment and skills: an empiricalstudy of French manufacturing. Cambridge Journal of Economics, 27:287-316.Greenan, N. and Guellec, D. (2000) Technological innovation and employment reallocation.Labour, 14, 4:547-590.Greenan, N., LHorty, Y and Mairesse J. (eds) (2002), Productivity, inequality and the digital economy.Cambridge, Mass., MIT Press.Hansen, A. H. (1964) Business cycles and national income, London, Allen and Unwin (expanded edn)Heertje, A. (1973) Economics and technical change. London. Weidenfeld and Nicolson.Howell, D. and Wolff, E. (1992) Technical change and the demand for skills by US industries,Cambridge Journal of Economics, 16, 128-146.Howell, D. (1996) Information technology, skill mismatch and the wage collapse: aperspective on the US experience. In OECD (1996b), 291-306.Institute for Prospective Technology Studies, European Science and Technology Observatory (IPTS-ESTO) (2001) Impact of technological and structural change on employment. Prospective analysis2020, Synthesis Report and Analytical Report, Seville, European Commission Joint Research Centre.Johnson, K. (1995) Productivity and unemployment: review of the evidence. In OECD, Investment,productivity and employment, Paris, OECD.Kalmbach P. and Kurz, H.D. (1990) Microelectronics and employment: a dynamic input-otput study ofthe West German economy, Structural Change and Economic Dynamics, 1, 317-86Karaomerlioglu, A. and Ansal, T. (2000), Compensation mechanisms in developing countries,in Vivarelli and Pianta (eds).Katsoulacos, Y. S. (1986) The employment effect of technical change, Brigthon, Wheatsheaf.Klette, T.J. and Forre, S.E. (1998). Innovation and Job Creation in a Small Open Economy:Evidence from Norwegian Manufacturing Plants 1982-92, Economics of Innovation and NewTechnology, vol. 5, 247-72Krueger, A. (1993) How computers have changed the wage structure: evidence from microdata 1984-1989, Quarterly Journal of Economics, 108, pp.33-60.Layard, R. and Nickell, S. (1985) The causes of British unemployment, National InstituteEconomic Review, 111, 62-85.Layard, R., Nickell, S. and Jackman, R. (1991), Unemployment: macroeconomic performanceand the labour market. Oxford, Oxford University Press.Layard, R., Nickell, S. and Jackman, R. (1994) The unemployment crisis. Oxford, OxfordUniversity Press.Leoni, R., Cristini, A. and Labory, S. (2003) Human resource management and firmsperformance. University of Bergamo discussion paperLeontief, W. and Duchin, F. (1986). The Future Impact of Automation on Workers. Oxford,Oxford University Press.Lundvall, B.A. (1987), Technological unemployment in a small open economy. In Lund, R.,Pedersen, P. and Schmidt-Sorensen, J. (eds), Studies in unemployment, Copenhagen, NewSocial Science Monographs, Institute of Organisation and Industrial Sociology. 22
  23. 23. Machin, S. (1996) Changes in the relative demand for skills. In Booth, A. and Snower, D.(eds) Acquiring skills. Cambridge, Cambridge University Press.Machin, S. and Van Reenen, J. (1998), Technology and Changes in Skill Structure: Evidencefrom Seven OECD Countries, Quarterly Journal of Economics, vol. 113, pp. 1215-1244.Machin, S. and Wadhwani, S. (1991) The effects of Unions on organisational change andemployment: evidence from WIRS, Economic Journal, 101, 324-30Matzner, E.R. and Wagner, M. (1990) The employment impact of new technology, Aldershot,Avebury.Meghir, C., Ryan, A. and Van Reenen, J. (1996) Job creation, technological innovation andadjustment costs: evidence from a panel of British firms. Annales d’economie et statistique,41-42: 256-273.Meyer-Krahmer F. (1992), The Effects of New Technologies on Employment, Economics ofInnovation and New Technology, vol. 2, 131-49.Mishel, L. and Bernstein, J. (1996) Technology and the wage structure: has technology’simpact accelerated since the 1970s?, Washington D.C., Economic Policy Institute.Nelson, R. and Phelps, E. (1966) Investment in humans, technological diffusion and economicgrowth. AEA Papers and Proceedings, 56:69-75.Nelson, R. and Winter, S. (1982), An Evolutionary Theory of Economic Change, Cambridge(Mass.), The Belknap Press of Harvard University Press.Nickell, S. and Kong, P. (1989). Technical Progress and Jobs, Centre for Labour Economics,Discussion Paper n.366, London, London School of EconomicsNoble, D. (1983). Present Tense Technology, Democracy. A Journal of Political Renewal and RadicalChange, vol. 3, no. 2 (pp. 8-24), no. 3 (pp. 70-82), no. 4 (pp. 71-93)Noble, D. (1984). Forces of Production: A Social History of Industrial Automation, New York, KnopfOECD (1994) The OECD Job Study. Evidence and explanations. Part I, Labour market trendsand underlying forces of change. Paris, OECD.OECD (1996a) Technology, productivity and job creation. Vol. 2 Analytical report, Paris,OECD.OECD (1996b) Employment and growth in the knowledge-based economy, Paris, OECD.OECD (2001), The new economy: beyond the hype. Paris, OECD.Padalino, S. and Vivarelli, M. (1997). The Employment Intensity of Economic Growth in theG-7 Countries, International Labour Review, 136: 191-213Pasinetti, L. (1981) Structural Change and Economic Growth, Cambridge, CambridgeUniversity Press.Perez, C. (1983), Structural change and the assimilation of new technologies in the economicand social systems, Futures, 15, 5: 357-75.Petit, P. (1995), Technology and employment: key questions in a context of highunemployment, Science Technology Industry Review, n.15: 45-63.Petit, P. (1995), Employment and Technological Change, in Stoneman, P. (ed), Handbook of theEconomics of Innovation and Technological Change, Amsterdam, North Holland, 366-408Petit, P. and Soete, L. (eds) (2001a), Technology and the future of European Employment.Cheltenham, Elgar.Petit, P. and Soete, L. (2001b), Technical Change and Employment Growth in Services: Analytical andPolicy Challenges, in Petit and Soete (eds)Pianta (2000) The employment impact of product and process innovation, in Vivarelli andPianta (eds)Pianta, M. (2001) Innovation, Demand and Employment, in Petit and Soete (eds.), 142-165Pianta, M. and Vivarelli, M. (1999) Employment dynamics and structural change in Europe,in Fagerberg, Guerrieri and Verspagen (eds)Pianta, M., Evangelista, R. and Perani, G. (1996) The dynamics of innovation andemployment: an international comparison, Science, Technology Industry Review, 18: 67-93. 23
  24. 24. Pini P. (1995), Economic growth, technological change and employment: empirical evidencefor a cumulative growth model with external causation for nine OECD countries, 1960-1990,Structural Change and Economic Dynamics, 6: 185-213.Pini, P. (1996), An Integrated Cumulative Growth Model: Empirical Evidence for NineOECD countries, 1960-1990, Labour, vol. 10, 93-150Pini, P. (1998), Occupazione, tecnologia e crescita: modelli interpretativi ed evidenzeempiriche a livello macroeconomico. Accademia nazionale dei Lincei, Sviluppo tecnologico edisoccupazione: trasformazione della società, atti del convegno, Roma,113-202.Piva M.C. and Vivarelli M. (2002) The skill bias: comparative evidence and an econometrictest. International Review of Applied Economics, 16, 3, 347-358Ricardo, D. (1951), Principles of Political Economy and Taxation, in Sraffa, P. (ed.), The Works andCorrespondence of David Ricardo, vol. I, Cambridge, Cambridge University Press (third edn 1821).Sanders, M. and ter Weel, B. (2000), Skill biased technical change: theoretical concepts, empiricalproblems and a survey of the evidence. Druid working paper, Copenhagen Business School andAalborg UniversitySchumpeter, J.A. (1934) Theory of Economic Development, Cambridge (Mass.), Harvard UniversityPress (1st edn 1911).Shaiken, H. (1984) Work transformed. Automation and labor in the computer age. New York,Holt, Rinehart and Winston.Simonetti, R., Taylor, K. and Vivarelli, M. (2000) Modelling the employment impact ofinnovation: do compensation mechanisms work? in Vivarelli and Pianta (eds)Simonetti, R. and Tancioni, M (2002). A macroeconometric model for the analysis of theimpact of technological change and trade on employment. Journal of InterdisciplinaryEconomics, 13:185-221.Sinclair, P. (1981) When will technical progress destroy jobs? Oxford Economic Papers, 31,1-18Smolny, W. (1998) Innovation, prices and employment: a theoretical model and anapplication for West German manufacturing firms, Journal of Industrial Economics, 46:359-81.Spezia, V. and Vivarelli, M. (2002), Innovation and employment: a critical survey. In N. Greenan, Y.LHorty and J. Mairesse (eds), 101-131Sylos Labini, P. (1969). Oligopoly and Technical Progress, Cambridge (Mass.), Harvard UniversityPress, first edn 1956Van Reenen, J. (1997). Employment and Technological Innovation: Evidence from U.K.Manufacturing Firms, Journal of Labor Economics, vol. 15, 255-84Van Reenen, J. (1996) The creation and capture of economic rents: wages and innovation in a panel ofUK companies. Quarterly Journal of Economics, 111, 443, 195-226Vivarelli M., Evangelista R., Pianta M. (1996), "Innovation and employment in the Italianmanufacturing industry", Research Policy, 25,1013-1026.Vivarelli, M. (1995). The Economics of Technology and Employment: Theory and Empirical Evidence,Aldershot, ElgarVivarelli, M. and Pianta, M. (eds) (2000), The Employment Impact of Innovation: Evidence and Policy,London, RoutledgeWhitley, J. D. and Wilson, R.A. (1982) Quantifying the employment effects of micro-electronics. Futures, 14,6:486-495.Wolff, E. (1996) Technology and the demand for skills. Science Technology Industry, 18:95-124. 24
  25. 25. Table 22.1A summary of approaches to Innovation and EmploymentMain research questions General approach Major streams of literature Key assumptions Main levelDoes technology creates or destroys jobs? and key findings and methodology of analysisEquilibrium of product and labour marketsWhat is the amount of jobs created/lost Labour economics Job demography and flexibility of labour markets Product and labour markets are in equilibrium Firms, industries,What is the skill composition Technical change favors more skilled workers, replaces The absolute level of job lost/gained is irrelevant macroeconomyWhat is the structure of wages the unskilled, exacerbates inequality Supply of educated workers shapes technical change Complementarity between ICTs and high skills Technological unemployment is irrelevantWhat are the returns from innovation Growth Technology, productivity, growth, employment: Standard production function, Industries, macro innovation may raise the natural rate of unemployment focus on process innovationWhat is the innovation input to growth New growth theory Endogenous innovation, growth and employment: Innovating and non innovating firms, spillovers, Macroeconomy unemployment may happen focus on process innovationDisequilibrium perspectivesWhat is the type of innovation Evolutionary Technological opportunities, variety, regimes: Innovation brings disequilibrium in markets Firms, industriesWhat is the amount/nature of unempl. firms strategies and industries outcomes are different New product markets emergeWhat are the structural factors Neo-Schumpeterian Techno-economic paradigms and long waves: Radical innovations, pervasiveness, diffusion Industries, macroWhat are the demand factors mismatches can lead to unemployment of new technology systems and ICTsWhat are the distribution effects Structural Sectoral composition of economies: Innovation is differentiated: constrasting IndustriesWhat are the institutions specificity of innovation and demand, different job results effects of new products and processes Regulationist Macro models for testing indirect effects of innovation: Industries are different, demand is important Macroeconomy compensation mechanisms may not work Countries are different, institutions are important 25
  26. 26. Table 22.2Presence, quality and impact of innovation in European countriesTechnological innovations introduced by firms in 1994-1996Percentage of all firms Innovating firms Manufacturing industry Firms introducing Firms introducing Firms introducing Percentage of turnover product process products new due to products Manufacturing Services innovations innovations to the market new to the firmEuropean Union 51 40 44 39 21 12(14 countries)Austria 67 55 60 49 24 13Belgium 34 13 31 22 14 6Denmark 71 30 58 51 27 7Finland 36 24 30 25 18 9France 43 31 38 31 20 9Germany 69 46 65 53 24 17Ireland 74 58 66 54 27 17Italy 48 31 37 41 26 8Luxembourg 42 49 32 29 21 naThe Netherlands 62 36 56 46 28 7Portugal 26 28 15 23 7 9Spain 29 na 24 25 11 10Sweden 54 32 48 38 25 14United Kingdom 59 40 52 37 19 8Norway 48 22 35 40 14 10Source: European Commission-Eurostat, Statistics on Innovation in Europe. Data 1996-1997, Luxembourg, 2001.Data from Community Innovation Survey 2 26
  27. 27. Table 22.3Effects of innovation on the quantity of employmentSelected empirical studiesStudy Countries Years Level of analysis Innovation data sources Results on employmentFirm level studiesEntorf and Pohlmeier, 1990 Germany 1984 Cross firm, manuf. Group of German firms Positive with product innovationMachin and Wadhwani, 1991 UK 1984 Cross firm, manuf. British workplace Positive industrial relations surveyBlanchflower, Millward and UK 1984 Cross firm, manuf. British workplace PositiveOswald, 1991 industrial relations surveyBrouwer, Kleinknecht and Netherlands 1983-1988 Cross firm, manuf. Dutch survey NegativeReijnen, 1993 Positive with product innovationMeghir, Ryan and UK 1976-1982 Panel of firms, manuf. SPRU innovation database PositiveVan Reenen, 1996 and patents more flexibilityKlette and Forre, 1998 Norway 1982-1992 Panel of manuf. firms Norway universe of manuf. NegativeVan Reenen, 1997 UK 1976-1982 Panel of manuf. firms Survey on UK firms PositiveSmolny, 1998 Germany 1980-92 Panel of manuf. firms Survey on German firms PositiveGreenan and Guellec, 2000 France 1986-1990 Cross firm, manuf. Innovation survey Positive at the firm level Cross sector Negative at the industry level for process innovations 27